Suchergebnisse

Argues for a rebalancing of survey researchers' emphasis on response rates & other simple measures of survey precision & quality with more attention to the profession's contributions to social & public policy decision making. This change in emphasis would make those who conduct surveys problem solvers as well as technicians. Emphasis on only what is easily measurable does affect how research is conducted. While technical knowledge & methodology are important, the overall goals of the project can be lost when response rates are blind to actual nonresponse error & bias, coverage errors, poor questions, poor study design, & failure to understand the issues. Problem identification & relevant research are a vital part of survey research. 12 References. L. A. Hoffman

An analysis of electoral behavior that deals with the limitations inherent in synchronic survey data methods. These limitations include lack of longitudinal perspective, overgeneralization from single, unique elections, & the conjoining of new methodologies with synchronic data. To promote comparative longitudinal research using survey data, causal models of the 1956, 1964, & 1972 presidential elections were developed. Methods used include a 7-variable causal model & cohort analysis. The 1956 model supports hypotheses emphasizing electoral stability. Party identification is the dominant influence on the vote, with issue preferences found to be weakly related to party identification or the vote. The 1964 & 1972 models show considerably greater voter responsiveness to political issues. Actual political & behavioral variability over time, rather than the electorate's "inherent limitations," or methodological inadequacies, are seen as responsible for previous divergent findings. 3 Tables, 3 Figures. Modified HA.

Tribute to Harry O'Neill who died 9/11/08.